ERIC Identifier: ED267435
Publication Date: 1986-00-00
Author: Purves, Alan C.
Source: ERIC Clearinghouse on
Reading and Communication Skills Urbana IL.
Testing in Literature. ERIC Digest.
Recently there has been increasing attention to testing of writing and
reading, but no parallel growth of interest in testing of literature. This is
unfortunate, not only because testing affects curriculum but also because
attention to the testing of literature can help teachers, English curriculum
specialists, test makers, and teacher educators refine their understanding of
literature and the way it is taught.
In order to examine the domain of literature study and to suggest ways by
which teachers might advance their own testing, one should look first at the
broad domain of literature and then at more specific objectives and school
outcomes. This process is necessary because literature curricula, even in
neighboring schools, may well have quite different philosophies, and testing of
student performance must take those differences into account.
WHERE ARE THE BROAD THRUSTS OF THE CURRICULUM IN LITERATURE?
In the United States, as around the world, there is no commonly shared view
about a single broad aim of instruction in literature, although there is general
consensus as to the range of major thrusts which differentiate one curriculum
from another. Although these thrusts have been given slightly different labels,
the definitions are stable: one literature curriculum focuses on texts and
knowledge of the literary and cultural heritage of a group; a second focuses on
the development of skilled readers and critics of literary texts; and a third
focuses on the encouragement of personal growth through reading and involvement
with the text (Dixon, 1966; Purves, 1978; Mandel, 1980). In any one school or
department, the actual curriculum may be a hybrid of two or even all three of
the thrusts, or the thrust may vary depending on the track or level of the
The implications of the three thrusts for testing are clear; no single test
can cover all three. The first calls for measures of recall, the second for
measures of skill, and the third for measures of attitude. Literature teachers
in a school system, therefore, should first define the broad goals of the
literature curriculum as it exists by examining the course of study, the
textbooks and other materials, and the kinds of assignments required of the
HOW CAN THE CONTENT AND OBJECTIVES OF THE LITERATURE CURRICULUM BE SPECIFIED?
One way of performing this examination of the curriculum is to set forth the
content of the curriculum and the kinds of behaviors or activities emphasized in
a grid (see the grid summary below). Various curricula would make different
applications of the principle of the grid and, therefore, different detailings
of the content and behavior. From any grid, the crucial next step is to
determine the emphasis of the curriculum for each cell. An historically based
curriculum would probably stress knowledge of specific texts, and contextual
(background) and cultural information. A more analytic curriculum would stress
application of critical terminology and expression of a preferred response.
Determining the relative emphasis of the various cells in the curriculum then
leads to the development of a set of test specifications.
A grid of content areas and behaviors might include these horizontal
headings: Behavior, Specific literary texts, Contextual information, Literary
theory, and Cultural behavior information. Vertical headings might consist of:
Be familiar with, Apply knowledge of specific literary texts to, Apply literary
history to, Apply literary theory to, Apply cultural information to, Respond to,
Express a pattern of preference for, Express a response to, Express a consistent
pattern of responses to, and Have positive attitudes and interests in
HOW ARE TEST QUESTIONS DEVELOPED?
Once the curriculum has been analyzed, and the various aspects of the domain
of literature study that are considered important in a school or class have been
determined, the next step is to develop a set of test specifications for that
domain. These specifications would include, for each cell to be measured, a
clear statement of what the cell means, perhaps with examples; a statement of
what type of measure (such as a true-false test, a series of essays, a
performance, or a set of out-of-class activities) would best let an observer
know what the students are doing with respect to that cell; and the criteria by
which adequate or superior performance would be determined.
Such specification would then lead to the development of tests,
questionnaires, and other measures that would indicate what the students have
gained. The purpose of such specification is primarily to insure the fairness of
the testing and the probability of comparable tests from semester to semester or
year to year. In many institutions, these test specifications are made available
to the students so that they are not surprised by the actual test or
performance, and they can better prepare.
These specifications and the resultant tests need to be examined carefully to
see whether the testing modes are appropriate to the emphasis in the cells. For
example, if the cell to be emphasized is that of expressing a response to a
piece of literature, it would make little sense to develop an elaborate
true-false or multiple-choice test; the emphasis of that cell is on expression,
perhaps written, perhaps oral, perhaps dramatic. Similarly, an out-of-class
essay on a play is not a good way of testing whether students can recall various
characters, scenes, and lines in the play.
In general, behaviors like recognition, recall, and even application are
efficiently tested through highly structured approaches such as true-false or
short answer. Behaviors that call for expression require more extensive measures
that ask students to classify, analyze, interpret, or evaluate texts. As a
general rule, the measures of these higher abilities should use texts that are
relatively unfamiliar to the student rather than texts extensively taught in
class. With the previously taught text, there is a greater possibility that the
student is recalling class discussion or notes rather than exercising a set of
analytic or interpretive skills.
HOW IS STUDENT PERFORMANCE JUDGED?
The final aspect of testing -- scoring or judging and then reporting student
performance -- is often the most difficult, particularly if more elaborate
measures are used. It is one thing to total up a number of right answers; quite
another to rate a piece of writing on the character of Stuart Little; still
another to judge individual performance in a dramatic interpretation of a Winnie
the Pooh story. In the latter cases, judgments are being made about the
rhetorical or dramatic skills of the students as well as their knowledge or
understanding of the text.
For the composition, many teachers seek to rate the quality of content and
the use of evidence and the reasoning, separate from rating the structure,
style, and mechanics of the writing. The latter aspects are important, to be
sure, but as far as performance in literature is concerned, it is the former
that counts. Similarly, in dramatic interpretation the students' understanding
of the characters and the dramatic situation should be rated separately from
their skills in speaking, gesturing, and the like. Research indicates that the
ratings of various aspects of performance are related to each other, but that
raters aware of the relationships can make distinctions between the content and
the form of a written or dramatic performance. For an overall grade in language
arts, of course, teachers might want to combine the two, but for the literature
aspect of the grade, the content is important.
Finally, some of the objectives of the literature curriculum are not subject
to judgment as to good or bad, but simply to description by the teacher. These
objectives concern attitudes about and interest in literature and literature
instruction. It is extremely important for these aspects to be measured through
some sort of questionnaire or informal interview, but the results might well
reflect less upon the student than upon the teacher and the curriculum.
Curriculum makers and teachers should comsider the possibility that a literature
course may produce students who are more knowledgeable but who do not want to
read literature, or the course may produce students who love literature but have
not learned anything. Ideally, the objectives in literature instruction should
form a harmonious whole.
FOR MORE INFORMATION
Dixon, John. GROWTH THROUGH ENGLISH. A REPORT BASED ON THE DARTMOUTH SEMINAR.
1966. ED 014 491.
Mandell, Barrett. THREE LANGUAGE-ARTS CURRICULUM MODELS: PRE-KINDERGARTEN
THROUGH COLLEGE. Urbana, Ill.: National Council of Teachers of English, 1980. ED
Purves, Alan. "Evaluation of Learning in Literature." In HANDBOOK ON
FORMATIVE AND SUMMATIVE EVALUATION OF STUDENT LEARNING, edited by B. S. Bloom,
J. T. Hastings, and G. Madaus. New York: McGraw-Hill, 1973.
Purves, Alan. "Evaluation of Learning in Literature." In EVALUATION IN
EDUCATION: AN INTERNATIONAL REVIEW SERIES, vol. 3. Oxford, England: Pergamon